Vostro 3560 (LA-8241P) dead short to ground

[spleenharvester]

Hi all.

I bought this system for other parts, but found the dead board is the rare AMD GPU version, so would be great to get it working again

I have unplugged everything, CPU, RAM daughterboard, the lot. Upon plugging in AC, the adapter immediately shorts. The system also seems to short a charged battery - no signs of life or voltages on any pin. There are no signs of water damage or any obviously fried components, and the short does not appear to be present when there is no attempt to plug in AC.

Any idea where to start with this one? I believe this should be the schematic? http://www.s-manuals.com/motherboard/compal_la-8241p

Thanks!

EDIT: I'll try and get some hi-res pics of the board up too if my camera plays nicely.

[spleenharvester]

Re: Vostro 3560 (LA-8241P) dead short to ground

Just found a bit of a strange mark on the GPU - I suspect it is some sort of burn mark and that the GPU is dead. Oh well.

EDIT: I've just scraped at it and it seems to be encrusted thermal paste, the core is actually intact. VGA_CORE_B+ (as measured at PJP800) is shorted to ground, however. Is it still worth trying with this board, or is the thermal paste likely burnt on, perhaps as a result of the GPU shorting?

[caspian]

Re: Vostro 3560 (LA-8241P) dead short to ground

A high-quality photo of the burn mark and the GPU would help.

[dellxps15]

Re: Vostro 3560 (LA-8241P) dead short to ground

if u have a lab psu is very simple to find the short...

[spleenharvester]

Re: Vostro 3560 (LA-8241P) dead short to ground

GPU looks completely clean now, was just thermal paste - I think possibly it was burnt on by heat from a GPU short but don't know for sure.

Upon further inspection, it seems the entire B+ rail has been shorted to ground. How would I go about using a lab PSU to find this short? Is there any way to do it with any other type of PSU or do I need to find one?

[dellxps15]

Re: Vostro 3560 (LA-8241P) dead short to ground

u need a lab psu to regulate ampere and voltages and ofcourse lab psu autoregulate max current by itself also. with it then find the short will take about 60 seconds.

[spleenharvester]

Re: Vostro 3560 (LA-8241P) dead short to ground

So I've just found out there is a 25V max, 5A max lab supply available at my university - what is the procedure for doing this, do I connect across the B+ rail to ground, start at 1V and feel for heat around the board? Given lots of things are connected to B+ do I risk damaging things above that voltage?

[spleenharvester]

Re: Vostro 3560 (LA-8241P) dead short to ground

So after reading up I've bought some freeze spray to try that trick - is the brief moment that the board is powered by a Dell AC adapter before its overcurrent trips likely to be enough to heat the faulty component, such that I can see it thaw without needing to try a benchtop PSU?

[piernov]

Re: Vostro 3560 (LA-8241P) dead short to ground

No, it won't cut it. You need to use a PSU for that. And yes, you should start at 1V, if ever a high-side MOSFET is shorted, the power could flow directly into the GPU or the CPU so check if they heat up with a low voltage or not. If nothing heats up you can go higher.

You could use an alkaline battery instead but that's not really convenient.

[dellxps15]

Re: Vostro 3560 (LA-8241P) dead short to ground

mainboard has dc jack plug soldered onboard or 4-5 wires with dc-jack separately ?

[spleenharvester]

Re: Vostro 3560 (LA-8241P) dead short to ground

Thanks guys - if I only need a few volts and <2A ish current to find the short, could I not theoretically just buy an adjustible converter like this for my own uses: https://www.ebay.co.uk/itm/UK-LM2596...V/172952085838 , add some sort of current limiter to the output, and just power it via a 9v battery?

Also the adapter input is the removeable, 2 red/2 black/1 white type that Dell sometimes uses.

[dellxps15]

Re: Vostro 3560 (LA-8241P) dead short to ground

for not burning a gpu-cpu and so on, check if gpu cpu coils are shorted. if not, u can put psu lab to 2Ampere and set it to 1 volt. ofcourse solder a wire to one red dc jack cable and put gnd on a metal plate. start psu lab.... if it doesnt absorb much current, psu led will be green. raise first ampere then voltages a bit until it get red. as soon it get red, let it run for 4-7 seconds and then power off psu. check wich part of mainboard is getting hot a bit. then power on again psu and start touching that zone till u find capacitor or mosfet getting hot as hell.
when u find one component hot, power off psu and desolder that component.after that, lower psu voltage back and increase it again to see if it get red again but not surpassing about 16-19 volt of B+ voltage . check back if without that component u have removed another one get hot.

if gpu was not shorted, u should not damage it....

remember 5Ampere and 4 volts means 20 watt and 20watt is a great amount of heat.... a gpu or bga can last longer with that power consumption, but a mosfet can damage board too, so dont take psu lab on red light too long. few seconds each time.

[spleenharvester]

Re: Vostro 3560 (LA-8241P) dead short to ground

Just checked the CPU/GPU coils (PL702, PL704 and so on) and none appear to be shorted. I will try this soon. Thanks!

[spleenharvester]

Re: Vostro 3560 (LA-8241P) dead short to ground

Until my freeze spray and bench PSU gets here, I tried hooking up a battery pack across B+ - I measure 4.5 volts and 2 amps going through the circuit when this happens, and the battery pack starts to heat up. However, I've had a cursory feel around with my hand and cannot find anything getting warm. Do I possibly need to increase the power, or is the problem perhaps a very small component that I'd only be able to spot with freeze spray?

[ktmmotocross]

Re: Vostro 3560 (LA-8241P) dead short to ground

take laptop 3.7V cell or 9v battery and inject it to charger circuit. battery have high amperage so be carefull not to burn bad component with traces. when connect this battery, some component start to burn

[spleenharvester]

Re: Vostro 3560 (LA-8241P) dead short to ground

take laptop 3.7V cell or 9v battery and inject it to charger circuit. battery have high amperage so be carefull not to burn bad component with traces. when connect this battery, some component start to burn

Thanks, have now hooked up a 9V battery (in circuit measures ~8 volts an 3A) and still nothing is heating up, even though the battery quickly gets hot (though I have to pull it within a few secs as the contacts get hot enough to burn skin).

It is also worth noting that there is not a short at the DC jack when there is no AC input - it reads 133k~ish, sometimes can't get a reading tho. The opposite side (pins 1,2,3 - drain?) of the immediate connecting MOSFET (PQ100, is this 'high side'?) is shorted, however, there is no short at pin 4.

[spleenharvester]

Re: Vostro 3560 (LA-8241P) dead short to ground

PQ100 (MDS3604) gets a bit warm when a 9V battery is placed between VIN and ground (measured amps in circuit = 2.5). Should this be replaced? I am a bit confused as to why it would be PQ100 when there is not a short on the VIN side.

(Also sorry for frequent posting - wish there was a longer edit timeout!)

[dellxps15]

Re: Vostro 3560 (LA-8241P) dead short to ground

pq100 got vin voltage.

as vin get in other place, measure pq100 pins if any of them is shorted, then remove pq100 and measure again if still pins are shorted.

[ktmmotocross]

Re: Vostro 3560 (LA-8241P) dead short to ground

P2 , P3 resistance to ground?

[spleenharvester]

Re: Vostro 3560 (LA-8241P) dead short to ground

P2 and P3 both measure at 1.5 ohms to ground (2.3 ohms minus 0.8 ohms of lead resistance). B+ is 0 ohms and VIN is 133 kiloohms.

I will resume troubleshooting this board in a few days when my equipment gets here. Does anyone know what the best way to replace small SMD components is without paying hundreds of £ for a high-end station (is my unregulated 20W iron likely to be too much power? Or is it the temperature regulation that matters?)